1 Differential expression analysis between brain and liver using EdgeR

In this analysis, coefficient 3 was used in the design matrix because it corresponds to the comparison of Liver vs. Brain. Brain is the reference level, thus, a positive logFC indicates higher expression in Liver, while a negative logFC indicates higher expression in Brain.

A total of 225 genes were found differentially expressed with an adjusted p-value (FDR) < 0.01 and an absolute logFC > 10. Specifically, 118 genes were overexpressed in Brain (compared to Liver) and 107 were overexpressed in Liver (compared to Brain). Most of these are protein-coding genes.

2 Gene ontology enrichment analysis (for biological processes) of the two sets of DEGs using GOstats

Brain-Specific Gene Expression

For the gene set overexpressed in Brain, 132 GO terms were enriched. Results were visualized with REVIGO:

Enriched terms include dominant categories related to developmental processes, neurogenesis, and regulation of gene expression. The enrichment of nucleic acid metabolism and DNA-templated transcription regulation suggests a high demand for gene expression control, essential for neuronal function and plasticity. The cell motility and neuron migration terms highlight key processes in brain development, as neurons must migrate to their correct locations during embryogenesis and synaptic remodeling. Additionally, the presence of multicellular organism development, cell fate commitment, and neurogenesis-related terms such as glial cell fate specification and spinal cord patterning, reinforces the brain’s structural and functional plasticity. The inclusion of epithelial cell differentiation and neural precursor cell proliferation suggests ongoing processes in maintaining neural stem cell populations and facilitating brain repair. The enrichment of dendrite self-avoidance and sensory organ morphogenesis highlighting the complexity of neuronal circuit formation and sensory processing.

Liver-Specific Gene Expression

For the gene set overexpressed in Liver, 119 GO terms were enriched. REVIGO visualization is shown below:

Key categories include lipid metabolism, peptide secretion, and response to external stimuli. The transport and localization of lipids, such as phospholipid efflux and vitamin transport, are crucial for hepatic function, as the liver plays a central role in lipid metabolism and systemic lipid homeostasis. The metabolic processes, including fatty acid metabolism and organic acid biosynthesis, highlighting the liver’s role in energy storage, detoxification, and synthesis of essential biomolecules. Additionally, the enrichment of peptide secretion and response to peptide hormone pathways suggests the liver’s endocrine functions, particularly in regulating glucose homeostasis through insulin and glucagon signaling. The presence of negative regulation of cytokine production and immune response processes indicates the liver’s involvement in immune modulation and inflammation control, essential for maintaining metabolic balance.

3 Differential splicing analysis using SUPPA between brain and liver

The table below summarizes the splicing events analyzed, including the absolute ΔPSI thresholds applied and the number of significant results obtained:

Splicing event ∆PSI threshold p-value threshold results (n)
Skipping exon (SE) > 0.5 < 0.05 11
Intron retention (RI) > 0.3 < 0.05 5
Mutually exclusive exons (MX) > 0.3 < 0.05 7
Alternative first exon (AF) > 0.7 < 0.05 3

Heatmaps were generated to visualize the top differential splicing events.

Skipping exon

Intron retention

Mutually exclusive exons

Alternative first exon

4 Analysis of H3K4me3 peaks using bedtools intersect

Peaks were classified as follows (and visualized using a bar plot):

  • Brain-specific peaks: 5,143
  • Liver-specific peaks: 7,324
  • Common peaks (shared between Brain and Liver): 19,356

5 Examples of differentially expressed genes with H3K4me3 peaks

A BED file was generated containing 200 bp upstream and downstream of each gene’s Transcription Start Site (TSS). This BED file was created to capture promoter regions where regulatory elements—like H3K4me3 peaks—are usually found (highlighted with blue in the images). Finally, DEGs were extracted and overlapped with the three sets of H3K4me3 peaks to correlate gene expression changes with chromatin state and chromatin accessibility.

UCSC Genome Browser extracts of 3 genes are shown below as examples:

Gene TSS overexpressed in brain with brain-specific peaks

Scg5 TSS exhibits brain differential expression. The brain-specific H3K4me3 peak at the TSS marks an active promoter, which correlates with higher chromatin accessibility as indicated by ATAC-seq peaks. This is consistent with the higher RNA-seq signal in the brain, confirming transcriptional activation. In contrast, the liver shows weak chromatin accessibility and transcriptional activity at this locus, despite the presence of H3K4me3, suggesting that additional tissue-specific regulatory elements contribute to Scg5 activation in the brain.

Gene TSS overexpressed in brain with common peaks

Diras1 locus exhibits strong transcriptional activity in brain while no activity in liver, as shown by RNA-seq data. A common H3K4me3 peak at the Diras1 promoter in both tissues suggests it is broadly marked as active. However, the chromatin accessibility profile (ATAC-seq) indicates that the promoter is more accessible in the brain, which likely facilitates higher transcription factor binding and gene activation.

Gene TSS overexpressed in liver with liver-specific peaks

The Spp2 TSS shows liver-specific ATAC-seq peak at the TSS, indicating open chromatin, coupled with a high H3K4me3 signal, marking an active promoter. This aligns with the strong RNA-seq expression in the liver, confirming active transcription. In contrast, the brain shows minimal chromatin accessibility, weak H3K4me3 enrichment, and negligible transcription as expected, as Spp2 is a liver differentially expressed gene.

6 Examples of alternative first exon

FAM214B gene

The FAM214B gene exhibits alternative first exon usage, as evidenced by RNA-seq data showing distinct transcription start sites (TSS) in liver and brain. ATAC-seq tracks confirm open chromatin regions at these TSS, while ChIP-seq data indicate differential histone modifications, reinforcing the tissue-specific regulation of transcription initiation.

HMBS gene

The Hmbs locus shows RNA-seq data revealing distinct transcription start sites. This differential exon usage is accompanied by variations in chromatin accessibility, showing strong peaks at the respective promoter regions in a tissue-specific manner. ChIP-seq data indicate distinct histone modification patterns, supporting tissue-specific regulation of Hmbs expression.